University of Birmingham

Potential for intracranial movements in pterosaurs Prondvai, Edina; si, Attila

Document Version Publisher's PDF, also known as Version of record Citation for published version (Harvard): Prondvai, E & si, A 2011, 'Potential for intracranial movements in pterosaurs', The Anatomical record.

Link to publication on Research at Birmingham portal

General rights Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes permitted by law.

•Users may freely distribute the URL that is used to identify this publication. •Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private study or non-commercial research. •User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?) •Users may not further distribute the material nor use it for the purposes of commercial gain.

Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document.

When citing, please reference the published version. Take down policy While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been uploaded in error or has been deemed to be commercially or otherwise sensitive.

If you believe that this is the case for this document, please contact [email protected] providing details and we will remove access to the work immediately and investigate.

Download date: 02. Oct. 2021 THE ANATOMICAL RECORD 294:813–830 (2011)

Potential for Intracranial Movements in Pterosaurs

1 † 2 EDINA PRONDVAI * AND ATTILA OSI 1Department of Paleontology, Eo¨tvo¨s Lora´nd University, Budapest, H-1117, Hungary 2Hungarian Academy of Sciences—Hungarian Natural History Museum, Research Group for Paleontology, Budapest, H-1083, Hungary

ABSTRACT Based on comparative anatomical, morphological, and phylogenetic considerations the potential of pterosaurs for cranial kinesis is assessed. Our investigation shows that whereas skeletally mature derived pterodac- tyloids have completely fused, rigid and doubtlessly akinetic skulls, skele- tally immature derived pterodactyloids and more basal pterosaurs possess key features in the morphology of their otic and basal joints that are suggestive of cranial kinesis, namely streptostyly. In addition, ptero- saurs exhibit an evolutionarily informative trend in the degree of cranial ossification, where it is low in most nonpterodactyloids (here named bife- nestratans), intermediate in Rhamphorhynchus and Archaeopterodacty- loidea, and high in derived pterodactyloids. Incomplete fusion could also indicate loose connections between skull elements. However, another cru- cial anatomical requirement of a kinetic skull, the permissive kinematic linkage is absent in all pterosaurian taxa. The fact, that the presence of permissive kinematic linkages in the skull is also a prerequisite of all types of cranial kinesis, provides hard evidence that all members of Pter- osauria had akinetic skulls. Thus, the presence of the morphological attributes indicative of intracranial movements in some pterosaurs must be explained on grounds other than real potential for cranial kinesis. It could either be of mechanical or ontogenetic importance, or both. Alterna- tively, it might be considered as the morphological remnant of a real, ki- netic skull possessed by the diapsid ancestors of pterosaurs. Anat Rec, 294:813–830, 2011. VC 2011 Wiley-Liss, Inc.

Key words: cranial kinesis; pterosaur; joint morphology; streptostyly

INTRODUCTION Additional Supporting Information may be found in the Kinesis and Pterosaurs online version of this article. The occurrence of cranial kinesis among a variety of This article was published online on 31 March 2011. Subse- quently, it was determined that the Acknowledgements were tetrapods have long been recognized (see Frazzetta, 1962 incomplete, and the correction was published on 14 April 2011. for references), but the conceptual definition was first Grant sponsors: Hungarian Scientific Research Fund, Bolyai provided by Versluys (1910, 1912) who defined a kinetic Fellowship (A.O¨ .), Hantken Foundation; Grant number: OTKA skull as allowing any intracranial movements between PD73021 (Paleo Contribution No. 128); Grant sponsor: Hungarian the elements excluded that of the lower jaw. Cranial Natural History Museum, Eo¨tvo¨s Lora´nd University Budapest. kinesis as a phenomenon, being present in the earliest *Correspondence to: Edina Prondvai, Department of Paleon- tetrapods (movable palatoquadrates and palatal and fa- tology, Eo¨tvo¨s Lora´nd University, Pa´zma´ny Pe´ter se´ta´ny 1/c, cial elements; Iordansky, 1989) and in some extant tele- Budapest H-1117, Hungary. Tel.: 0036 20 526 00 60. osts and amphibians (Rieppel, 1978; Summers and E-mail: [email protected] Wake, 2005), is generally considered a plesiomorphic Received 23 July 2010; Accepted 1 December 2010 character (Iordansky, 1990) and is thought to be most DOI 10.1002/ar.21376 prominent within archosaurs and lepidosaurs (Herrel Published online 31 March 2011 in Wiley Online Library et al., 1999). However, clear evidence of true cranial (wileyonlinelibrary.com).

VC 2011 WILEY-LISS, INC. 814 PRONDVAI AND O†SI

Fig. 1. Three levels (A, B, C) of diapsid phylogeny gradually focus- among archosaurian clades focusing on the conspicuously different ing on the archosaurian clades (C). A, Interrelation of the major diapsid position of Pterosauria in the two cases (WST, Weighted Supertree; clades (after Lee, 2001); B, division of archosauriformes (after Sereno, SM, Supermatrix Tree). Groups of special significance for the EPB 1991); C, two possible outcomes for the phylogenetic relationships evaluation of this study are boldfaced. kinesis, which should be distinguished from slight move- (for further information see Supporting Information), ments occurring at many patent sutures and allowing was the only form of kinesis ever suggested for ptero- dissipation of mechanical stresses, exists unambiguously saurs. Most pterosaurologists have regarded the ptero- only in some squamates and birds among amniotes (Hol- saurian skull as universally akinetic (e.g., Wellnhofer, liday and Witmer, 2008). 1978; Buffetaut et al., 2002; Fastnacht, 2005). Nevertheless, some forms of cranial kinesis have been In the light of the dominance of more derived ptero- and probably will be suggested for numerous extinct ver- saurs with firmly fused skull bones in the fossil record, tebrates, particularly dinosaurs. In contrast to dino- this attitude is easy to understand. On the other hand, saurs, the notion of intracranial movements in another based on the apparent, although sometimes debated archosauromorph group, the pterosaurs is not so com- close affinities of pterosaurs to dinosaurs (Hone and mon. Except for the work of two authors, Arthaber Benton, 2008, and see Fig. 1 for the position of Pterosau- (1919) and Wild (1978, 1984), who regarded the Early ria in a broader phylogenetic context), for which cranial Jurassic Dorygnathus banthensis and the Upper Triassic kinesis has been proposed on several occasions (e.g., Col- Eudimorphodon ranzii, respectively, as having streptos- bert and Russell, 1969; Galton, 1974; Norman, 1984; tylic quadrate, and Bennett (1996a) who used the term Norman and Weishampel, 1985; Chiappe et al., 1998; ‘‘metakinetic skull’’ as a streptostylic character sugges- Mazzetta et al., 1998, see Supporting Information), and tive of archosauromorph nature of pterosaurs in his phy- on certain morphological attributes of some pterosaurian logenetic analysis, thus accepting Wild’s (1978, 1984) skulls it seems reasonable to pay more attention to the concept for Eudimorphodon, this issue has largely been potential of intracranial movements in pterosaurs. ignored. Hence, streptostyly, which refers to the antero- For the acquirement of the necessary theoretical back- posterior rotation of the quadrate about the otic joint ground, Supporting Information is provided which CRANIAL KINESIS IN PTEROSAURS 815 contains detailed information on basic concepts such as bony lamellae (e.g., craniofacial hinge in prokinesis the different forms, morphological correlates, functional of birds, see Supporting Information). If one of these significance, occurrence, origin, and evolution of cranial features is present, similarly to synovial joints, it can kinesis in other diapsid reptiles. In this Supporting In- assure mobility in the critical otic and basal regions of formation, the extremely kinetic skull of Serpentes is the skull. not regarded. The third criterion which has been cited by Holliday and Witmer (2008) as another morphological correlate of Osteological Aspects of Kinesis in Extant and cranial kinesis is the state of development of the protrac- tor muscles in the skull. They considered the presence of Extinct Taxa well-developed preotic and levator pendants as osteologi- To reveal features that are suggestive of cranial kine- cal indicators of protractor muscles (e.g., m. protractor sis several methods have been in use mainly in extant pterygoideus) which could have operated powered intra- taxa (Frazzetta, 1962, 1983; Smith and Hylander, 1985; cranial movements (Holliday and Witmer, 2008). The Patchell and Shine, 1986; Condon, 1987; Herring and presence, reconstructed size, and attachment areas of Teng, 2000; Metzger, 2002; see Supporting Information). most muscles in fossil groups are, however, obscure at Among the most important features is the presence of best, and the functional significance of the protractor morphological correlates which include the co-operating muscle group in extant taxa is sometimes also inconclu- muscle, connective, and skeletal tissues assuring the sive (Gussekloo and Bout, 2005). proper functioning of the involved joint systems, and The fourth criterion is referred to as permissive kine- which can be assigned to certain types of kinesis (Bahl, matic linkage which includes those taxon-specific fea- 1937; Bu¨ hler, 1981; Rieppel and Gronowski, 1981; Zusi, tures that permit observable intracranial movements in 1984; Rieppel, 1993; Arnold, 1998; Metzger, 2002; see extant taxa possessing true kinetic skull. These are in Supporting Information). Nevertheless, it must be general related to elimination or mobility-modification of remembered that the absence of these morphological cor- bony elements surrounding the movement centers that relates can allow the exclusion of cranial kinesis, but the would otherwise hinder the intracranial movements. presence of them can only indicate the potential for Furthermore, Holliday and Witmer (2008) defined movement and cannot definitively prove its presence in three categories of inferred kinetic state, of which, obvi- vivo, thus they must be viewed with a measure of cau- ously, only the first two criteria can be applied to extinct tion (Throckmorton, 1976; Herrel and De Vree, 1999; forms: Metzger, 2002; Holliday and Witmer, 2008). When it comes to extinct vertebrates, significant 1. partially kinetically competent: the skull possesses amount of information is lost due to the incompleteness key synovial joints and protractor muscles but lacks or lack of preservation of soft tissues that must have bony gaps permitting movement had important role in intracranial movements. The only 2. fully kinetically competent: the skull possesses key sy- available data in most cases are the osteological fea- novial joints and protractor muscles as well as per- tures. Holliday and Witmer (2008) have defined four cri- missive bony linkages but lacks demonstrable teria or morphological correlates that are indispensible movement in vivo concerning inferences of powered cranial kinesis in fossil 3. kinetic: the skull possesses key synovial joints, pro- taxa. tractor muscles and permissive bony linkages as well The first two criteria regard the detectable presence as demonstrable movement in vivo. of mobile joints in the otic (quadratosquamosal) and ba- sal (basipterygopterygoid) regions of the skull (Holliday Thus, extinct taxa, such as dinosaurs or pterosaurs and Witmer, 2008). They stated that the mobile joint may at most be fully kinetically competent. In extant type in these regions must be synovial. ‘‘Synovial joint,’’ diapsids, true cranial kinesis only occurs along with the the presence of which is often referred to as one of the reduction of certain cranial elements (lower temporal most important criteria in kinetic bony connections, bar in lepidosaurs but also the postorbital bar in gekko- implies a noninterdigitate, finished, smooth joint with tans and varanids; the supratemporal, postorbital, and synovial capsule (Holliday and Witmer, 2008). The lacrimal bar in birds, see Supporting Information), so it osteological correlates of synovial joints are (1) the seems parsimonious to infer that some form of reduction presence of convex and complementary concave joining of bones might also be necessary to acquire a kinetic surfaces of the participating elements; (2) the smooth diapsid skull. articular surface indicative of hyaline cartilage cover- ing; (3) rough, parallel striated zone and occasionally large pits distal to the smooth surface revealing the Skulls of Pterosaurs in General presence of the joint capsule and ligament attachments, respectively (Holliday and Witmer, 2008). However, To discuss the issue of cranial kinesis in pterosaurs, a skull elements which are to form a mobile joint (mobil- brief general description of pterosaurian skulls is neces- ity at least to the extent over which it can be referred sary. The skull of pterosaurs is generally lightly built to as kinetic joint), can also be connected in different (fenestrated) and elongated, and they always have com- ways, for instance via ligament (e.g., quadratopterygoid plete lower temporal arch formed by mainly the jugal ligament), or in some special cases movement can occur and partially the quadratojugal. even in a smooth/slightly interdigitating fibrous joint Until recently there had been basically two morpho- (e.g., frontal-parietal joint in mesokinesis of geckoes, logical types of pterosaurs distinguished, the more basal see Supporting Information). Furthermore, flexibility nonpterodactyloid (generally referred to as ‘‘rhampho- can be ensured via inbuilt flexion zones formed by thin rhynchoid’’) and the more derived pterodactyloid 816 PRONDVAI AND O†SI

Fig. 2. Line drawing of the two skull-morphotypes based on the p, parietal; pl, palatine; pmx, premaxilla; po, postorbital; pt, pterygoid; genera A, Rhamphorhynchus and B, Anhanguera representing the q, quadrate; qj, quadratojugal; sq, squamosal; 1, naris; 2, antorbital fe- bifenestratan and monofenestratan morphotype, respectively. Note the nestra; 3, orbit; 4, supratemporal fenestra; 5, lateral temporal fenestra; length of the rostrum, the state of the naris (1) and antorbital fenestra 6, choana; 7, suborbital fenestra; 8, pterygo-ectopterygoid fenestra; 9, (2), and the state of two palatal fenestrae (8, 9) as main differences subtemporal fenestra; 10, interpterygoid vacuity; 11, cranioquadrate between the two basic ‘‘bauplans.’’ Abbreviations: bpt, basipterygoid; opening. ec, ectopterygoid; f, frontal; j, jugal; l, lacrimal; mx, maxilla; n, nasal;

constructions, which have significant differences in their lent in this study with the former nonpterodactyloid– body architecture (Fig. 2). In contrast to ‘‘nonpterodacty- pterodactyloid concept. Nevertheless, the terms ‘‘bifenes- loids’’ the term ‘‘pterodactyloids’’ implies not only a mor- tratan’’ and ‘‘monofenestratan’’ are used here only in a phology-based category but forms a valid monophyletic morphotype-sense and do not imply real phylogenetic group, as well. Two new pterosaur genera discovered in categories. China, Darwinopterus and Wukongopterus (Lu¨ et al., The main differences between the two skull morpho- 2009 and Wang et al. 2009, respectively), however, have types can be summarized as follows: whereas bifenestra- challenged the concept of this morphology-based distinc- tans (Fig. 2A) have a naris (1) and an antorbital tion by having a mixture of nonpterodactyloid and ptero- fenestra (2) separated by a bony bar consisting of the dactyloid characters in their skeleton. Accordingly, a conjunction of the maxillary process of the nasal and the modified concept is needed to define different pterosau- nasal process of the maxilla, monofenestratans (Fig. 2B) rian morphotypes. Since the main object of this study is usually have more elongate rostrum with a confluent the skull, a new terminology was suggested on one hand and very large nasoantorbital fenestra (1þ2). In addi- by M. Witton (pers. com.) on the other hand by the tion, the confluence of certain palatal fenestrae (8þ9, authors of this study for distinguishing pterosaurian Fig. 2B) is also a characteristic of monofenestratans (O†si skull-constructions without regarding the phylogenetic et al., 2010). The basic ‘‘bauplan’’ of the skull of the two position of the taxa concerned. This new, morphology- morphotypes is illustrated in Fig. 2. Some monofenestra- based concept is defined by virtue of the relation tans are edentulous, while all hitherto known bifenestra- between the naris and antorbital fenestra: if they are tan pterosaurs have teeth and sometimes elaborate separated, the skull is referred to as bifenestratan;if dentition. they are confluent, the applied term for the skull is Phylogenetic interrelationships of pterosaur genera monofenestratan (suggested by M. Witton). The bifenes- (based on Dalla Vecchia, 2009, and Andres and Ji, 2008) tratan and monofenestratan skull-morphotypes corre- without regarding the new problematic taxa, Darwinop- spond to the former nonpterodactyloid and pterodactyloid terus and Wukongopterus are shown in Fig. 3. þ Darwinopterus (and possibly Wukongopterus, as well; In this article, we investigate the morphological corre- M. Witton, pers. com.) constructions, respectively. How- lates of potential intracranial movements in pterosaurian ever, since Darwinopterus and Wukongopterus for practi- skulls using the comparative strategy and evaluation cal reasons (bad preservation and limited access) are not applied by Holliday and Witmer (2008) for dinosaurs, and included in the current investigation, the taxonomical consider the results in the context of what we recently composition of the two new morphotype-groups is equiva- know about the phenomenon of cranial kinesis. CRANIAL KINESIS IN PTEROSAURS 817 Frankfurt, Germany; SMNK, Staatliches Museum fu¨r Naturkunde, Karlsruhe, Germany; SMNS, Staatliches Museum fu¨ r Naturkunde, Stuttgart, Germany; TMM, Texas Memorial Museum, University of Texas, USA; WDC, Wyoming Dinosaur Center, Wyoming, USA; YPM, Peabody Museum of Natural History, Yale University, New haven, USA; ZMNH, Zhejiang Museum of Natural History, Zhejiang Province, China.

MATERIALS AND METHODS In this study, pterosaurs are subjected for the first time to an investigation on their potential or incapacity for cranial kinesis. To infer the likelihood of cranial kine- sis in pterosaurs in a phylogenetic context, we applied Extant Phylogenetic Bracket (EPB) method (Witmer, 1995), since cranial kinesis involves unpreserved soft tis- sues to a high degree. Because of different interpreta- tions of the phylogenetic position of pterosaurs, the two most widely accepted approaches have been used in the evaluation; (1) Pterosauria is the sister group to Dino- sauria (Hone and Benton, 2008, Fig. 1C, WST) being bracketed by birds and crocodiles (Fig. 1B); (2) Pterosau- ria is a basal Archosauromorpha (Bennett, 1996a) being bracketed by lepidosauromorhs and archosauromorphs (Fig. 1C, SM). The evaluation was based on the search for morpho- logical correlates indicative of cranial kinesis and was carried out in 27 different pterosaur genera most of Fig. 3. Pterosaur phylogeny based on the recent cladistical analy- which had either almost complete skull material or cru- ses. A, Interrelationships of bifenestratan pterosaur species with indi- cial skull elements of good preservation. Of the 27 inves- cation of their relation to the more derived Pterodactyloidea (modified tigated genera, 13 can be assigned to bifenestratans and from Dalla Vecchia, 2009); B, interrelationships of the monophyletic 14 to pterodactyloid monofenestratans. Among the 71 Pterodactyloidea on family and generic level (modified from Andres investigated specimens, 45 were available for personal and Ji, 2008). examination (see Table 1). Twenty-six of the referred original specimens, which were not attainable for the authors, were assessed based on the related literature, Institutional Abbreviations casts, and published photos of high resolution (see Table AMNH, American Museum of Natural History, New 2). Where it was possible, all joints of the skull elements York, USA; BMNH, British Museum of Natural History, which might be relevant or might refer to any type of in- London, England; BNM, Bu¨ ndner Naturmuseum, Chur, tracranial movements have been evaluated. Braincases, Switzerland; BSPG, Bayerische Staatsammlung fu¨ r Pal- as expected to be fused in all adult specimens and puta- a¨ontologie und Geologie, Mu¨ nchen, Germany; BXGM, tive circumorbital elements are not considered in the Benxi Geological Museum, Liaoning Province, China; morphological investigation. Skeletally immature speci- CD, Desire´e Collection of Rainer Alexander von Blitters- mens have also been examined to identify possible dorff Rio de Janeiro, Brasil; CM, Carnegie Museum of changes in the biomechanical behavior of the skull dur- Natural History, Pittsburgh, USA; DNPM MCT, Museu ing ontogeny. Ciencias da Terra, Setor de Paleontologia do Departa- The phylogenetic tree of Dalla Vecchia (2009) and that mento Nacional de Producao Mineral, Rio de Janeiro, of Andres and Ji (2008) was used in the evaluation of Brasil; GMV, Chinese Geological Museum, Beijing, results in a phylogenetic context. However, it must be China; GPIUB, Geologisch-Pala¨ontologisches Institut, taken into account that the tree of Dalla Vecchia (2009) Universita¨t Bonn; IGO, Instituto de Geologı´a y Paleonto- differs from the results of most other phylogenetic analy- logı´a, La Habana, Cuba; IVPP, Institute of Vertebrate ses in the position of anurognathids, which group is con- Paleontology and Paleoanthropology, Beijing, China; sidered a derived clade by Dalla Vecchia (2009) and a KUVP, Museum of Natural History, University of Kan- basal nonpterodactyloid group by Bennett (1996a), Kell- sas, Kansas, USA; MCSNB, Museo Civico di Scienze ner (2003), and Unwin (2003). Naturali di Bergamo, Italy; MN, Museu Nacional, Rio de Janeiro, Brazil; MPUM, Dipartimento of Scienze della RESULTS Terra dell’Universita` di Milano, Italy; MSFN, Museo EPB Evaluation Friulano di Storia Naturale, Udine, Italy; NSM, National Science Museum, Tokyo, Japan; PIN, Palaeon- The EPB method (Witmer, 1995) can indeed be very tological Institute, Russian Academy of Sciences, useful, since it provides the most parsimonious assump- Moscow, Russia; RGM, Naturalis (Nationaal Natuurhis- tion to infer the likelihood of an unknown feature in an torisch Museum), Leiden, The Netherlands; SMNF, extant taxon. However, the reliability of EPB decreases Senckenberg Forschungsinstitut und Naturmuseum, significantly with increasing uncertainty of the 818 PRONDVAI AND O†SI TABLE 1. Examined specimens that were available for personal investigation with indication of their generally accepted or presumable ontogenetic stage by ‘‘i,’’ immature (juvenile or subadult); ‘‘m,’’ mature (adult) and ‘‘?,’’ uncertain Taxa available for personal investigation Inventor number of the specimen Feasible ontogenetic stage Eudimorphodon ranzii MCSNB 2888 m Carniadactylus sp. MPUM 6009 m (‘‘Eudimorphodon ranzi’’ sensu Wild, 1978) Caviramus filisurensis BNM 14524 m (‘‘Raeticodactylus filisurensis’’ sensu Stecher, 2008) Dimorphodon macronyx BMNH 41212-13 m BMNH R 1035 m Dorygnathus banthensis BSPG 1938 I 49 m SMNS 18969 m SMNS 50164 m SMNS 50914 m SMNS 51827 m SMNS 55886 m WDC-CTG-001 m Rhamphorhynchus muensteri BSPG AS VI 34 m BSPG 1867 II 2 m BSPG 1889 XI 1 i BSPG 1927 I 36 m BSPG 1929 I 69 m BSPG 1934 I 36 i BSPG 1938 I 503 i BSPG 1955 I 28 m BSPG 1989 XI 1 m SMNK PAL 6596 ? SMNS 52338 m SMNS 56980 m Campylognathoides liasicus SMNS 18879 m SMNS 50735 m Campylognathoides zitteli SMNS 9787 m Scaphognathus crassirostris SMNS 59395 i Austriadactylus cristatus SMNS 56342 m Pterodactylus antiquus BSPG AS I 739 m Pterodactylus kochi BSPG AS V 29 m (P. antiquus sensu Bennett, BSPG AS XIX 3 m 1996b) BSPG 1878 VI 1 i BSPG 1883 XVI 1 m BSPG 1937 I 18 m SMNF R 4072 m Pterodactylus micronyx BSPG 1971 I 17 i (Gnathosaurus subulatus sensu Bennett, 1996b) Pterodactylus sp. BSPG 1936 I 50 i SMNF R 4074 m Germanodactylus cristatus BSPG 1892 IV 1 m Ctenochasma gracile BSPG 1920 I 57 m (Ctenochasma elegans sensu Bennett, 2007b) Ctenochasma sp. SMNS 81803 m Anhanguera sp. SMNK uncatalogued m Araripesaurus santanae BSPG 1982. I. 90 i (Santanadactylus araripensis sensu Bennett, 1993) Tapejara wellnhoferi SMNK PAL 1137 i phylogenetic relationships of the taxon in question. Fur- phylogenetic position can result in all three levels of in- thermore, the higher taxonomic level the bracketing ference depending on which group is considered within taxa represent, the higher the variability in the exam- the extant components. Applying Bennett’s (1996a) ined trait can be within the bracketing taxa themselves. approach, the two ways of bracketing pterosaurs already In pterosaurs, this problem is even worse due to their give three different levels of inference. If the bracket is ambiguous phylogenetic position and the large distance formed by lepidosauromorhs and archosauromorphs, between them and their presumed closest living rela- both of which represent very high taxonomic levels, we tives. The two main interpretations concerning their can find kinetic, as well as akinetic representatives in CRANIAL KINESIS IN PTEROSAURS 819 TABLE 2. Examined specimens that were evaluated on the basis of related literature, casts and published photos of high resolution Taxa unattainable for personal Inventor number of Feasible ontogenetic investigation Investigation based on the specimen stage Carniadactylus rosenfeldi Dalla Vecchia, 2009 MSFN 1797 m Campylognathoides liasicus Wellnhofer, 1974 CM 11424 m Scaphognathus crassirostris cast SMNS 80203 GPIUB 1304 m Cacibupteryx caribensis Gasparini et al., 2004 IGO-V 208 m Batrachognathus volans Ryabinin, 1948; Dalla PIN 52-2 m Vecchia, 2002 Dendrorhynchoides curvidentatus Dalla Vecchia, 2002 GMV2128 m Anurognathus ammoni Bennett, 2007a; high uncatalouged m resolution photos by H.Tischlinger Rhamphorhynchus muensteri high resolution photos CM 11434 m by B. Mueller Gegepterus changi Wang et al., 2007 IVPP V 11981 i Anhanguera bittersdorffi Campos and Kellner, 1985 MN 4805-V m Anhanguera piscator Kellner and Tomida, 2000 NSM-PV 19892 i Anhanguera sp. Kellner, 1996 MCT 1501-R i Coloborhynchus spielbergi Veldmeijer et al., 2006 RGM 41880 m Istiodactylus latidens Hooley, 1913 BMNH R 0176 ? Istiodactylus sinensis Andres and Ji, 2006 NGMC 99-07-011 m Tapejara wellnhoferi Wellnhofer and Kellner, 1991 AMNH 24440 ? Kellner, 1989 CD-R-080 i Kellner, 1996 MCT 1500-R i Sinopterus dongi Wang and Zhou, 2003 IVPP V13363 ? Huaxiapterus benxiensis Lu¨ et al., 2007 BXGM V0011 ? Pteranodon sp Bennett, 2001 KUVP 976 m KUVP 2212 m YPM 1177 m Zhejiangopterus linhaiensis Cai and Wei, 1994 ZMNH M1330 m Quetzalcoatlus sp. Kellner and Langston, 1996 TMM 41961-1 m TMM 41954-62 m The generally accepted or presumable ontogenetic stage of the specimens is indicated by ‘‘i,’’ immature (juvenile or suba- dult); ‘‘m,’’ mature (adult) and ‘‘?,’’ uncertain. both bracketing clades. Choosing Rhynchocephalia Joint morphology. In pterosaurian skulls, two within Lepidosauromorpha for one extant component characteristic types of joints can occur: fibrous joints and Suchia within Archosauromorpha for the other gives (syndesmoses) and synovial joints (see Fig. 4A—D, and Level III inference (decisive negative assessment) for E,F, respectively). Thin portion of bony elements, which cranial kinesis in pterosaurs, since all extant species of might be capable of flexion like the bending zones in Rhynchocephalia (Sphenodon) and Suchia (Crocodilia) birds, may also be present. have akinetic skulls. However, using Squamata as one Fibrous joint can further be subdivided into two cate- and Aves as the other component, in both clades of gories based on the articular surface morphology of the which intracranial movements are characteristic, cranial connecting elements. The first and most common fibrous kinesis in pterosaurs is Level I inference (decisive posi- joint is formed by overlapping bony processes (Fig. 4A,B) tive assessment). Applying the same extant clades but in of the participating, mostly viscerocranial elements (see different combinations, e.g., Squamates—Suchia and Table 3). This overlapping arrangement, which broadly Rhynchocephalia—Aves will both imply Level II infer- speaking corresponds to the term scarf joint or sutura ence (equivocal assessment) with one component exhibit- squamosa (Szenta´gothai and Re´thelyi, 2006), mostly ing cranial kinesis and the other possessing akinetic implies broad, oblique contact areas between the bones skull. Using the phylogenetic interpretation of Hone and that were joined by fibrous connective tissue. Table 3 Benton (2008), EPB gives again a Level II inference of summarizes the connecting skull elements that show cranial kinesis in pterosaurs bracketing pterosaurs this joint morphology and the overlapping-overlain rela- between akinetic crocodiles and kinetic birds. Thus use tions of them detected in almost all referred specimens, of the EPB in predicting the kinetic potential of ptero- where the preservation and incomplete fusion state saur skulls is inconclusive. allowed proper identification of joint morphology. The occurrence of overlapping fibrous joints between the referred skull elements is consistent and apparently uni- versal among pterosaurs. The nature of connection Morphological Correlates between some palatal elements is ambiguous, because, The morphological observations are subdivided into apart from the few cases where significant information three main categories, namely joint morphology of skull can be gained from specimens that are either exposed elements, osteological correlates of protractor muscles, in palatal view or three dimensionally preserved and ossification degree of the skull as a unit. (see O†si et al., 2010), the palatal construction of most 820 PRONDVAI AND O†SI

Fig. 4. Forms of joints occurring in the pterosaurian skull repre- fibrous fusion of elements demonstrated by the ventral aspect of the sented here by Dorygnathus banthensis SMNS 55886 (A, B), WDC- frontal-parietal connections; E and F, apparently synovial joints dem- CTG-001 (C, D, F) and SMNS 51827 (E) specimens without consider- onstrated by the distinct articular surfaces of the basipterygoid proc- ing complete fusion of elements. A, overlapping fibrous joint demon- esses (E) and the cephalic and mandibular condyles of the quadrate strated by the postorbital-jugal connection in which the ascending (F). Black arrows point to the articulating areas. Abbreviations: bpt, process of jugal overlaps the descending process of postorbital; B, basipterygoid; bs, basisphenoid; cc, cephalic condyle; f, frontal; fli, overlapping fibrous joint demonstrated by the lacrimal-jugal, jugal- impression of the frontal lobes of the brain; ij, unfused fibrous joint maxilla, and nasal-maxilla connections in which the processes of jugal, with interdigitating suture; j, jugal; l, lacrimal; mc, mandibular con- maxilla, and nasal overlap the lacrimal, jugal, and maxilla, respectively; dyles; mx, maxilla; n, nasal; oj, overlapping fibrous joint; p, parietal; pj, C, patent suture preceding interdigitating fusion state demonstrated unfused fibrous joint with patent suture; pmx, premaxilla; po, postorbi- by the dorsal aspect of contralateral frontals, parietals and the fronto- tal; pt,pterygoid; q, quadrate; sj, synovial joint. Scale bar: 1 cm. parietal connections (dorsal view); D, interdigitating suture preceding

TABLE 3. Skull elements forming overlapping Dorygnathus banthensis SMNS 50702 and Pterodactylus joints with one another in the pterosaurian skull sp. BSPG 1936 I 50. The same could have applied to the palatine-maxilla junction by virtue of the morphology of Overlying element Overlain element their contacting areas suggested by O†si et al. (2010) and pmx n, f indeed seems to be the case in Dorygnathus banthensis j mx, po, qj BSPG 1938 I 49, SMNS 50914, WDC-CTG-001, Rham- nl,mxphorhynchus muensteri SMNK PAL 6596, Pterodactylus ljsp. BSPG 1936 I 50. The ectopterygoid-maxilla junction po f, sq qj q is also overlapping in Rhamphorhynchus muensteri CM pl mx 11434. sq p The second fibrous joint category is the patent or interdigitating suture where the articulating parts of Abbreviations: f, frontal; j, jugal; l, lacrimal; mx, maxilla; n, the elements are much more robust and distinct and do nasal; p, parietal; pmx, premaxilla; po, postorbital; pt, pter- not become as thin as the overlapping bony processes ygoid; q, quadrate; qj, quadratojugal; sq, squamosal. (Fig. 4C,D). Patent (open) suture (sutura plana, Szenta´- gothai and Re´thelyi, 2006, and see Fig. 4C) generally pterosaurian skulls is hypothetical at best. However, refers to an earlier developmental state of the interdigi- based on its long, narrow wing-like processes, the ptery- tating suture (Fig. 4D) and can be identified by the goid is expected to have connected to the ectopterygoid, straight, noninterdigitating contact areas in skeletally palatine, and maxilla via overlapping fibrous joint. This immature specimens (Fig. 4C). These suture morpholo- situation is found between the pterygoid and palatine in gies were present between the counter-elements of the CRANIAL KINESIS IN PTEROSAURS 821

Fig. 5. Examples for quadrates having distinct cephalic condyle the proximal view of cephalic condyle in Dorygnathus banthensis (cc) at the proximal end of the ascending shaft which is indicative of WDC-CTG-001 where the fine texture of the surface indicative of hya- synovial quadrate-squamosal (otic) joint. A, Campylognathoides zitteli line cartilage covering becomes apparent. Abbreviations: cc, cephalic SMNS 9787; B, Campylognathoides liassicus SMNS 18879; C, Arari- condyle; q, quadrate; sqc, cotyle on squamosal. Scale bare of A, B, pesaurus santanae BSP 1982. I. 90; D, Tapejara wellnhoferi SMNK C, D: 1 cm; E: 0.5 cm. PAL 1137; E, Dimorphodon macronyx BMNH 41212-13; F, close up of frontal and parietal bones, in the frontoparietal suture, specimen, the suture between the premaxillae is still between the counterparts of the premaxillae, the pre- visible on the dorsal surface, but there is no sign of it on maxillae and nasals, and between some palatal elements the ventral surface. This might suggest an earlier fusion such as the quadratopterygoid joint. Patent sutures of the ventral or medial sides of the skull elements. Dis- were present between the counterparts of the frontals in tinguishing overlapping and interdigitating suture mor- Dorygnathus banthensis SMNS 18969, 50164, WDC- phology has proven to be difficult if the joint is CTG-001 (Fig. 4C), Anurognathus ammoni (unca- completely fused. talouged), Rhamphorhynchus muensteri BSPG 1938 I Synovial joints are not frequently found in the skull of 503, Pterodactylus antiquus BSPG AS I 739, Pterodacty- pterosaurs; however, there are some taxa and/or ontoge- lus kochi BSPG 1878 VI 1, Anhanguera piscator NSM- netic stages in which there is a seemingly synovial con- PV 19892, between the counterparts of the parietals in nection between the quadrate and squamosal (Figs. 4F, Dorygnathus banthensis SMNS 50164, WDC-CTG-001 5) and the basipterygoid and pterygoid bones (Figs. 4E, (Fig. 4C), Campylognathoides liasicus SMNS 18879, 6A). Among the investigated specimens, the bifenestra- Anurognathus ammoni (uncatalouged), and in the fron- tans Eudimorphodon ranzii MCSNB 2888, Austriadacty- toparietal joint in Dorygnathus banthensis SMNS 18969, lus cristatus SMNS 56342, Carniadactylus rosenfeldi WDC-CTG-001 (Fig. 4C), Anurognathus ammoni (unca- MSFN 1797, Carniadactylus sp. MPUM 6009, Cavira- talouged), Rhamphorhynchus muensteri BSPG 1938 I mus filisurensis BNM 14524; Dimorphodon macronyx 503, Campylognathoides zitteli SMNS 9787, Campylog- BMNH 41212-13 (Fig. 5E); Dorygnathus banthensis, nathoides liasicus SMNS 18879, Scaphognathus crassir- WDC-CTG-001 (Fig. 7F), SMNS 18969, 50164, 55886, ostris GPIUB 1304, Pterodactylus kochi BSPG 1878 VI Campylognathoides liasicus SMNS 18879, Campylogna- 1. The frontoparietal suture is often marked by a well- thoides zitteli SMNS 9787 (Fig. 5A,B), Scaphognathus developed transversal ridge even in skeletally immature crassirostris GPIUB 1304, Rhamphorhyynchus muen- specimens with patent suture (e.g., Dorygnathus ban- steri BSPG 1938 I 503, and pterodactyloids Araripesau- thensis, WDC-CTG-001, Fig. 4C). Other specimens rus santanae, BSPG 1982. I. 90 (Fig. 5C), Anhanguera where the preservation and incomplete fusion state piscator NSM-PV 19892, Tapejara wellnhoferi SMNK allowed the recognition of fibrous joints between the PAL 1137 (Fig. 5D), AMNH 24440 most likely have had referred skull elements showed interdigitating sutures synovial quadrate-squamosal, i.e., otic joint. In these often with very faint suture lines (e.g., frontoparietal specimens, the quadrate has a distinct, well-developed joint in Rhamphorhynchus muensteri 1934 I 36, Campy- condylus cephalicus on its ascending process (Fig. 5) lognathoides liasicus SMNS 50735, Scaphognathus cras- which fits in the corresponding cotylus on the ventral sirostris SMNS 59395, Pterodactylus kochi SMNF R side of the squamosal (Fig. 5B). The surface texture of 4074, Araripesaurus santanae BSPG 1982 I 90, etc.). condylus cephalicus is superbly preserved in Dorygna- Different fusion state on the dorsal and ventral side of thus banthensis WDC-CTG-001 and undoubtedly indica- the same elements also occurred; e.g., in Dorygnathus tive of hyaline cartilage covering (Fig. 5F). The nature of banthensis WDC-CTG-001 on the dorsal surface, the the basipterygoid-pterygoid, i.e., basal joint is ostensibly frontoparietal suture is apparently still open (Fig. 4C), synovial in Eudimorphodon ranzii MCSNB 2888, Car- whereas on the ventral surface it already shows an niadactylus rosenfeldi MSFN 1797, Cacibupteryx cari- interdigitating appearance (Fig. 4D). In the very same bensis IGO-V 208, Dorygnathus banthensis SMNS 822 PRONDVAI AND O†SI

Fig. 6. Joint types of three articulating palatal bones, basipterygoid an apparently synovial joint with the pterygoid (basal joint). B, Fused (bpt), pterygoid (pt) and quadrate (q). A, The superbly preserved quadrate-pterygoid unit lying isolated on the slab of Campylogna- Rhamphorhynchus muensteri CM 11434 (‘‘Carnegie specimen’’) with in thoides liasicus SMNS 50735. Black and white arrows indicate the ap- situ arrangement of these elements clearly shows that whereas the proximate or clear joining areas of the bones. Scale bar: 1 cm. quadrate is fused to the pterygoid, the basipterygoid process forms

Fig. 7. Comparison of the ossification degree of the skull of A,a of Anhanguera (A), the different skull elements of Dorygnathus (B) can derived pterodactyloid, Anhanguera sp. (SMNK uncatalogued) and B, easily be recognized due to the distinct suture lines (indicated by a bifenestratan, Dorygnathus banthensis (SMNS 55886). Note that white arrows) that trace out the individual bone shapes. whereas there are no visible suture lines between the skull elements

18969, 50702, 51827 (Fig. 4E), Scaphognathus crassir- As for possible ‘‘bending zones,’’ almost all viscerocra- ostris GPIUB 1304, Rhamphorhynchus muensteri BSPG nial elements of the pterosaurian skulls are considerably 1989 XI 1, SMNK PAL 6596, CM 11434 (Fig. 6A). In thin, and many of them are even lamella-like. However, these specimens, the basipterygoid processes have dis- on the medial surface they are always reinforced tinct, blunt ending which articulate with the correspond- mechanically by a bracing system of thickened bony ing concave facet of the pterygoid (Fig. 6A). The spars. For example in Dorygnathus banthensis WDC- expected rough surface structure distal to the presum- CTG-001, the medial surface of the jugal is mechanically ably synovial articular surfaces (on the quadrate and strengthened by a tetraradiate bracing system that run basipterygoid) which would be suggestive of the presence along the long axis of the four processes, whereas the of a synovial capsule is hardly discernible in any of the maxilla is supported medially by a vertical lamina that specimens either due to preservational or preparational enhances both the lateral nasal process and the ventral artifacts. Acidic preparation for instance can result in palatal plate (O†si et al., 2010). destruction or modification of the original bone surface (e.g., Dorygnathus banthensis WDC-CTG-001), whereas Protractor muscles. Following the criterion on transparent coating used to protect the fossils from the presence and development of protractor musculature chemical or mechanical effects might disguise important defined by Holliday and Witmer (2008), the orbitotempo- morphological attributes (e.g., Campylognathoides liasi- ral part of the braincase has been investigated to search cus SMNS 18879, Dorygnathus banthensis SMNS for possible attachment areas of musculus levator ptery- 18969). Nevertheless, both the basal and otic joints are goidei (mLPt) and musculus protractor pterygoidei apparently synovial fulfilling two of the defined morpho- (mPPt). These two muscle groups are generally consid- logical criteria of cranial kinesis (Holliday and Witmer, ered as being of crucial importance in active kinesis: 2008). mLPt and mPPt are constrictor dorsalis muscles and CRANIAL KINESIS IN PTEROSAURS 823 play role in protraction of the kinetic system (the quad- trally to contact the pterygoids via short but slightly rate in the streptostylic movement), whereas the adduc- expanded basipterygoid processes (Bennett, 2001). tor musculature, m. adductor mandibulae externus and In Coloborhynchus spielbergi RGM 401880, the basi- m. pterygoideus retract it (Herrel et al., 1999; Metzger, sphenoid narrows posteriorly, where it contacts the 2002; Holliday and Witmer, 2007, 2008). Although the basioccipital. attachment areas can vary among different extant taxa, The sutures of the laterosphenoid with the surround- by and large mLPt originates on the ventral side of the ing elements are mostly unclear, thus there are different parietal or on the fused laterosphenoid-prootic complex interpretations concerning its extent in pterosaurs. Ben- and inserts on the dorsal surface of the pterygoids; nett (2001) considered it the element forming the inter- mPPt originates on the basisphenoid and/or prootic and orbital septum, whereas Kellner (1996) regarded the inserts on the dorso-medial side of the pterygoid (Herrel interorbital septum as a separate element (pseudome- et al., 1999; Metzger, 2002; Holliday and Witmer, 2007, sethmoid). Due to these (and probably also interspecific) 2008). differences, the contacts of the laterosphenoid are not The morphology of the pterygoid is well known in dif- alike in the two interpretations. According to Bennett ferent pterosaurian taxa (for morphological description (2001) in Pteranodon, it has a Y-shaped cross section see O†si et al., 2010). The corpus of this tetraradiate ele- dorsally, where it contacts the frontals, and ventrally it ment is generally thin compared with its processes in develops into a strut-meshwork reaching down to the bifenestratans (e.g., Dorygnathus banthensis SMNS basisphenoid. Anteriorly, its dorsal margin extends into 50164, Campylognathoides liasicus SMNS 50735, Rham- the median pneumatic space. It has also lateral proc- phorhynchus muensteri CM 11434) but might be more esses that extend from the anterodorsal corner and con- robust in monofenestratans (e.g., in Anhanguera sp. tact the lacrimals. The processes are not fused to the SMNK uncatalogued) without lateral process (see O†si lacrimals but have blunt terminations. Posteriorly, the et al., 2010). In case protractor muscles were to attach laterosphenoid may overlap the prootic and opisthotic. on its surface, they probably would have inserted on the In contrast, Kellner (1996) described the laterosphenoid dorsal side of its corpus or near its posterior end which as being connected anteriorly to the frontal and via a was more robust and connected to the basipterygoid and medially directed process to the ‘‘pseudomesethmoid,’’ quadrate. and posteriorly to the parietal and prootic in Tapejara The orbitotemporal region of the pterosaurian skulls, wellnhoferi MCT 1500-R. In Anhanguera sp. MCT 1501- however, is poorly known since the bones surrounding R, the laterosphenoid is expanded under the ventrolat- the endocranial cavity are pneumatic, thus badly eral surface of the parietal and contacts the prootic post- crushed in most cases (Bennett, 2001). In this respect, eroventrally (Kellner, 1996). there are only handful examinable specimens, which The prootic along with the opisthotic and basioccipital bear reliable information for the adequate reconstruction form the walls and floor of the endocranial cavity. In of this skull region. These mostly three dimensional, Pteranodon, the prootic and opisthotic together form the well-prepared specimens (Tapejara wellnhoferi MCT otic capsule (Bennett, 2001). The prootic has a complex 1500-R, AMNH 24440, Anhanguera sp. MCT 1501-R, morphology in Tapejara wellnhoferi MCT 1500-R, and Pteranodon sp. KUVP 976, 2212, YPM 1177, Coloborhyn- lies between the laterosphenoid, parietal, and opisthotic chus spielbergi RGM 401880) were not accessible for the in Anhanguera sp. MCT 1501-R (Kellner, 1996). authors of the recent article, thus the relevant braincase All three braincase elements are pierced by foramina elements namely the basisphenoid, laterosphenoid, and forming the passages of different cranial nerves. Ele- prootic have been evaluated based exclusively on litera- vated areas at the base of the laterosphenoid-prootic ture data (Kellner, 1996; Wellnhofer and Kellner, 1991; complex identified as the attachment areas for mLPt Bennett, 2001; Veldmeijer et al., 2006). Detailed photos and mPPt in dinosaurs (Holliday and Witmer, 2008) or and explanatory figures of this region are found in Kell- muscle scars are not reported in any of the referred ner (1996), Wellnhofer and Kellner (1991), and Bennett specimens, nor there is any other detailed study investi- (2001). gating skull musculature other than those which The basisphenoid in pterosaurs is mostly elongated manipulate the mandible (Fastnacht, 2005; O†si, 2010). and anteroventrally directed in basal nonpterodactyloids Owing to the extensive free surface on the braincase and as well as in the more derived pterodactyloids (Kellner, pterygoids, however, there are still remaining areas for 1996). Posteriorly, it contacts the basioccipital to which potential protractor muscles to attach. it is firmly fused with very faint or obliterated suture line, whereas anteriorly it forms the basipterygoid proc- esses which articulate with the pterygoids. It borders Degree of skull ossification. In the evaluation the cranioquadrate opening medially, and its anterior of fusion state of the skull, skeletally mature and imma- end forms the posterior margin of the interpterygoid va- ture specimens must be distinguished and ideally only cuity (Bennett, 2001). In Tapejara wellnhoferi MCT mature specimens should be taken into account. Our 1500-R, the dorsal part of the basisphenoid is expanded investigation implies that most bifenestratans and but ventrally becomes thin. Anteriorly, it is connected to derived monofenestratans (corresponding to Dsungarip- the interorbital septum by numerous bony struts, but its teroidea in phylogenetic context [Kellner, 2003]) seem to base is free of these trabeculae (Kellner, 1996); the same be consistently distinct in this regard (Fig. 7) with the condition that was found in Pteranodon (Bennett, 2001). derived, bifenestratan genus Rhamphorhynchus and ba- In AMNH 24440, the basisphenoid-parasphenoid sal monofenestratans (corresponding to Archaeoptero- complex is an expanded, somewhat concave bony plate dactyloidea in phylogenetic context [Kellner, 2003]) (Wellnhofer and Kellner, 1991). The basisphenoid of occupying a fusion state somewhere between the two Pteranodon is an elongate element extending anteroven- extremities. Most cranial elements of those bifenestratan 824 PRONDVAI AND O†SI

Fig. 8. The occurrence of different morphological correlates of been taken from related literature. Taxa without symbols have not potential cranial kinesis (see A, for symbol legend) in B, bifenestratan been investigated here. Note that all bifenestratans of well-known skull pterosaurs represented at genus level and C, pterodactyloid ptero- morphology except for Rhamphorhynchus have a quadrate with dis- saurs represented at genus and family level demonstrated in a phylo- tinct cephalic condyle and possess an incompletely fused skull as genetic context (modified from B, Dalla Vecchia, 2009, and C, Andres adult. Low ossification degree of the skull is also characteristic of ba- and Ji, 2008). Those taxa which have been examined by the authors sal pterodactyloids (Archaeopterodactyloidea). personally are marked by asterisks, while data for the remainder have CRANIAL KINESIS IN PTEROSAURS 825 pterosaur specimens that are generally considered adults the morphological correlates of neither typical nor sliding are distinct (Eudimorphodon ranzii MCSNB 2888, synovial joints (see above). The morphology and struc- Campylognathoides liasicus CM 11424, Carniadactylus tural arrangement of the overlapping joints in the ptero- sp. MPUM 6009, Caviramus filisurensis BNM 14524, saurian skulls are rather indicative of syndesmoses or Dimorphodon macronyx BMNH 41212-13, Dorygnathus fibrous joints; something similar to the rigid scarf joints banthensis SMNS 50702, 51827, 55886 [Fig. 7B], Austria- found in crocodiles (O†si et al., 2010). In addition, the pro- dactylus cristatus SMNS 56342, Anurognathus ammoni, posed sliding motion would have also been hampered by uncatalouged), very often disarticulated or even scattered the immobility of adjacent bones. In agreement with Hol- (Batrachognathus volans PIN 52-2, Dendrorhynchoides liday and Witmer (2008) here, we reject the variety of in- curvidentatus GMV2128, Dimorphodon macronyx BMNH tracranial movements via such ‘‘sliding’’ joints supposed R 1035, Dorygnathus banthensis SMNS 18969, 50164, for dinosaurs; thus the joints in the pterosaurian skulls 50914, BSPG 1938 I 49, WDC-CTG-001, Campylogna- in which the participating elements uniformly show this thoides liasicus SMNS 18879, 50735, Campylognathoides arrangement are regarded as being incapable of any sig- zitteli SMNS 9787). In contrast, those specimens of nificant movement. Hence, all the joints listed in Table 3 derived monofenestratan pterodactyloids which are con- have been excluded from having potential for kinesis in sidered adults have completely ossified skulls with very pterosaurs. faint or no suture lines (Anhanguera bittersdorffi unca- Patent (Fig. 4C) or interdigitating fibrous joints (Fig. talouged holotype, Anhanguera sp. SMNK uncatalogued 4D) which are mostly completely fused without any [Fig. 7A], Pteranodon sp. KUVP 976, 2212, YPM 1177, traceable suture line in skeletally mature specimens Tapejara wellnhoferi CD-R-080, Zhejiangopterus lin- indicate complete immobility between the connecting ele- haiensis ZMNH M1330; Huaxiapterus benxiensis BXGM ments. For example in Dorygnathus banthensis SMNS V0011 Quetzalcoatlus sp. TMM 41961-1, 41954-62) 50164, the parietals are fused to the frontals, but the (Fig. 7A). An intermediate degree of skull ossification contralateral elements are not, which implies that these seems to be exhibited in almost all specimens of the two elements fuse earlier during ontogeny than either derived bifenestratan Rhamphorhynchus muensteri the contralateral frontals or parietals to each other (in (BSPG AS VI 34, 1867 II 2, 1927 I 36, 1929 I 69, 1934 I case there was a determinate order of fusion of cranial 36, 1955 I 28, 1989 XI 1, CM 11434, SMNK PAL 6596) elements). This refers to the immobile nature of the and in monofenestratan archaeopterodactyloids (Ptero- frontoparietal joint already in earlier ontogenetic stage. dactylus antiquus BSPG AS I 739, Pterodactylus kochi However, in Campylognathoides liasicus SMNS 50735 BSPG AS XIX 3, 1937 I 18, SMNF R 4072, R 4074, Ger- the frontoparietal suture seems to be still open, whereas manodactylus cristatus BSPG 1892 IV 1, Ctenochasma the contralateral elements are already fused without gracile BSPG 1920 I 57, Ctenochasma sp. SMNS 81803). suture impression. Thus, either there is interspecific Figure 8 represents the summary of results, where the variation in the order of fusion of cranial elements or distribution of different morphological correlates, which there is no determinate fusion order whatever. In any might suggest potential for intracranial movements in case, the transversal ridge indicating the frontoparietal pterosaurian skulls, is indicated in a phylogenetic suture in every ontogenetic stage would have prevented context. the mesokinetic dorsoventral rotation of the frontal along this suture, thus precluding the potential of the pterosaurian skull for mesokinesis (see Supporting Infor- DISCUSSION mation). Based on the phylogenetic position of ptero- Among the morphological features of pterosaurian saurs within Archosauromorpha (Fig. 1B,C after Sereno, skulls assessed in this study, there are some which seem 1991; and Hone and Benton, 2008, respectively), the ab- to indicate intracranial movements, and others which sence of mesokinetic movements at the frontoparietal apparently do not allow any mobility between the joint is to be expected. Fibrous joints found in other referred elements. areas of the skull are also rigid and immobile. The quadrate-squamosal joint with distinct condylus cephalicus on the quadrate (Fig. 5) that has a surface Joint Morphology and Mobility texture characteristic of hyaline cartilage covering (Fig. The most common joint type, the overlapping joint (Fig. 5F) and with corresponding cotyle on the squamosal 4A,B) strongly resembles those found in dinosaurian (Fig. 5B) suggests the presence of a synovial otic joint. skulls and has been considered by several authors to be The articular morphology found between the basiptery- capable of sliding motion (see Supporting Information). goid and pterygoid indicates the presence of synovial However, as Holliday and Witmer (2008) pointed out, joint, too. With respect to function, it might seem there is no extant equivalent of such moveable joints; in obvious to presume that with such synovial morphology fact this joint arrangement must have prevented any the quadrate must have been capable of anteroposterior kind of motion between the joining elements. Neverthe- rotation (streptostyly) along with the pro- and retraction less, there are synovial joints in which the participating of the basal unit via synovial basipterygoid-pterygoid elements do not show a typical convex condyle–concave joint. However, there is a serious problem with this cotyle morphology but rather both articular ends are assertion. This problem is related to other morphological straight, well defined and robust structures with smooth constraints with which all concerned skull elements surfaces that can slide alongside each other like e.g., in must be in accordance for the animal to achieve strep- the basipterygoid-pterygoid joint in the palate of birds tostyly. The animal either has to reduce some bony ele- (Zusi, 1993) or Varanus (pers. obs.). The overlapping ments lateral and medial to the distal end of the joints with elongated, tapering bony processes found in quadrate or its connections to these bones must be sig- the pterosaurian and dinosaurian skulls, however, exhibit nificantly mobile (e.g., synovial or ligamentous 826 PRONDVAI AND O†SI connection) for it to be capable of anteroposterior rota- structurally possible. However, even if the internal con- tion. In addition, if the basal joint was to be functional, struction of the element itself had allowed bending in too (pro- and retraction of the muzzle coupled with quad- the thinner zones (e.g., the nasal does not have stiffen- rate movement), the basal unit as well as the skull roof ing system), the arrangement of adjacent bones would must have contained flexible regions. No reduction or not have allowed any bending movement. For instance, mobility modification can be observed in the adjacent if there was any movement allowed along the thin region bones and skull regions in pterosaurs. The bones lateral of the nasal, it would be a laterodorsal rotation relative to the quadrate form typical overlapping joints (quadra- to the dorsal process of the premaxilla. However, this tojugal overlaps the quadrate and is overlain by the ju- motion must have been impeded by the adjacent but gal, see Table 3), and these bones are connected to all surely fixed bones (maxilla, lacrimal, frontal) to which it surrounding elements via overlapping joint, thus the was connected via fibrous, immobile joints. Thus the position of them can be considered fixed. Hence they exceptionally thin bone walls, which are so characteristic form a most probably immobile, rigid lower temporal of the whole skeleton of pterosaurs, must have contrib- arch which, along with the fused quadrate-pterygoid uted to the lightness of the skull as well as postcranium joint, would not allow the anteroposterior movement of rather than ensured structural flexibility. the quadrate. In addition to that, the squamosal has a ventral process that overlaps the ascending process of the quadrate laterally which further fixes the position of Protractor Muscles and Mobility the quadrate. Accordingly streptostyly can almost cer- tainly be ruled out even in taxa such as Eudimorphodon The reconstruction of muscles in extinct animals is am- ranzii (contrary to O†si, 2010), Dorygnathus banthensis, biguous at best. Reconstructions inferred from the pres- Tapejara wellnhoferi (juvenile), etc., where the form and ence of muscle scars, extensive bony surfaces, tubercles, construction of the quadrate-squamosal and basiptery- etc., and/or from muscle arrangement found in the closest goid-pterygoid regions are indicative of kinesis. In spite extant relatives (EPB) can be misleading for many of their synovial morphology, the basal and otic joints reasons. were not to form movable joints, and this quasi-contra- First, not all muscles leave muscle scars on their diction must be resolved. The term ‘‘synovial’’ is a struc- attachment areas. Muscle scars are to be expected in tural joint category, which implies the morphology of the those muscles which connect to the bone via collagenous connecting elements but does not necessarily refer to a tendon, whereas muscles with fleshy attachment on the mobile (diarthrodial) joint (Holliday and Witmer, 2008). bone do not necessarily leave traces of their origination Since many cranial elements, including the quadrate, os- or insertion areas. On the other hand, surface modifica- sify endrochondrally, i.e., hyaline cartilage is present tion on bones can be caused by other tissues, such as during the ossification process (Dixon, 1997), this articu- glands, neurovascular bundles, etc. (Witmer, 1995), thus lar surface structure can simply imply that these ele- not necessarily by muscles. Even if the muscle most ments were connected by fibro- or hyaline cartilage thus probably have had tendinous contact to the bone (con- forming a cartilaginous rather than real synovial joint. cluded from EPB), the state of preservation, preparatio- Cartilaginous joints or synchondroses are typically found nal artifact or the relatively hidden position of the bone between basicranial bones (connected by hyaline carti- (e.g., laterosphenoid, prootic) can all prevent the detec- lage) and may ossify with age, or they form the interver- tion of muscle scars. tebral discs (hyaline þ fibrous cartilage) (Szenta´gothai Second, the attachment areas need not be very exten- and Re´thelyi, 2006). Although the condyle-cotyle mor- sive or distinct in any other way. In fact, free surface phology is not usually found in cartilaginous joints, area on the bone, be it ever so limited, might provide there is a significant variability and transition in the places of attachment for smaller muscles. morphology of cartilaginous joints (Szenta´gothai and Third, inferring from closest living relatives might be Re´thelyi, 2006), thus initially both of these types could inherently dangerous in two ways. In the case of ptero- have accounted for the described morphology in the ba- saurs depending on the interpretation of their phyloge- sal and otic joints of pterosaurs. Nevertheless, if carti- netic position (see for different phylogenies in Bennett, lage was present in these regions only to facilitate bone 1996a; Hone and Benton, 2008), the bracketing taxa in growth during earlier ontogenetic stages as it has al- EPB might well be crocodiles, birds or lepidosauro- ready been suggested for the cranial bones of dinosaurs morphs. These clades are very distinct morphologically (Holliday and Witmer, 2008), this morphology would as well as functionally (Schwarz et al., 2007) bearing rel- only be present in skeletally immature specimens. Yet atively few common features and sometimes homology- some bifenestratan specimens which show this feature relations in their skeleton and musculature are still not are considered adults, thus skeletally mature (e.g., Eudi- clear (Holliday and Witmer, 2007). In addition, interspe- morphodon ranzii MCSNB 2888, Scaphognathus crassir- cific differences in the musculoskeletal system might be ostris GPIUB 1304, Campylognathoides liasicus SMNS very high. For instance, regarding the skull of birds, 18879, Dorygnathus banthensis SMNS 50702, etc., see some taxa exhibit passive cranial kinesis without any above). Hence, either the ontogenetic age of these speci- protractor muscle activity (rhynchokinesis in palaeogna- mens has to be re-defined or the cartilaginous covering, thous birds, Gusseklo and Bout, 2005, see Supporting which adults also possessed, had a role other than sim- Information), while others (e.g., toucan) have completely ply providing places of bone growth. This leads us fur- akinetic skulls (for reference see Zusi, 1993). Moreover, ther to the problem of identifying ontogenetic stages in in the lepidosaur Sphenodon there is even intraspecific fossils and is discussed below. variance in the presence of the pro- and retractor The thin lamella-like elements, indeed, might have muscles of the pterygoids, despite that they have com- been relatively flexible and bending could have been pletely akinetic skull (Metzger, 2002). CRANIAL KINESIS IN PTEROSAURS 827 In sum, the probability that pterosaurs possessed the Stress-strain distribution and forces acting on a skull protractor muscle system necessary for streptostyly can- during feeding can be quite different when comparing a not be estimated; their role in cranial kinesis, however, rigid, united skull (derived monofenestratans) with an based on the evaluation of joint morphology, can be incompletely fused and thus structurally more elastic excluded. skull (basal bifenestratan). In his biomechanical investi- gation, Fastnacht (2005) demonstrated that ‘‘butt-ended sutures resist compression but will fail in tension; scarf Degree of Ossification and Mobility joints with oblique articulating surfaces accommodate The apparent incomplete ossification of the skull in tensile and compressive forces in all directions with most bifenestratan pterosaurs such as Eudimorphodon, usually only minor movement [ :::]. Interdigitating con- Dorygnathus, Campylognathoides, etc., and the interme- tacts resist compressive and tensile forces and prevent diate fusion state found in Rhamphorhynchus and slipping between adjacent bones’’ (pp. 31-32). He fur- archaeopterodactyloids like Pterodactylus or Cteno- thermore stated that most derived pterosaurs have chasma, seem to occur irrespective of ontogenetic age and fused skull with no sutural impression, which he suggest loose connection between the cranial elements. referred to as ‘‘single unit skull.’’ In contrast, the Upper However, concerning fossil vertebrates, the terms adult, Triassic and Lower Jurassic pterosaurs possess a skull juvenile, skeletally mature, or immature often lead to con- that consists of a cluster of bones that disarticulate fusion either because there is no consensus in their use or post mortem, thus their cranium is a composite rather we do not know enough of the ontogeny of the extinct or- than a single unit skull (pp. 32). He also suggested that ganism to define exactly what is meant by which (see Ben- in brevirostrine taxa (e.g., anurognathids) the incom- nett, 1995, 1996b for a review). The small number of plete fusion of the skull elements may be a direct me- known specimens in most taxa further prevents the estab- chanical consequence of the high strains present in lishment of a more systematic terminology. relatively short skulls (Fastnacht, 2005, pp. 185). In Although it seems to be true that most known bifenes- this context, the issue of feeding strategy is of crucial tratan pterosaurs and archeopterodactyloids have incom- importance: there could have been a principal difference pletely ossified skulls with at least distinct sutures, it in feeding behavior between basal and derived ptero- still might be possible that all of these hitherto known saurs which was responsible for this alteration in ossifi- specimens are skeletally immature. The continuous cation degree. Further evidence for different feeding debate on whether the establishment of a new species is mechanics stems from their probably different jaw mus- reasonable or the specimen only represents a different culature; an inference derived on one hand from the ontogenetic status of an already described species also complexity of the basal pterosaur palates compared illustrates this quandary (see Bennett, 1996b, 2007b; with those seen in some pterodactyloids (M. Witton, Dalla Vecchia, 2002, 2009). Those cases where the rest pers. com.), on the other from the difference in the in- of the skeleton seems to be mature and only the skull clination degree of the quadrate and the maximum indicates immaturity (e.g., Dorygnathus banthensis gape (O†si, 2010). WDC-CTG-001, with fused syncarpals but isolated skull Nevertheless, this field also abounds in uncertainties; bones), raise an issue on whether there is a determined hence it is not the most useful aspect to settle this ques- sequence of ossification process among different skeletal tion. Still, due to the aforementioned morphological fea- elements with the skull ossifying last. tures, the incomplete skull ossification of the considered On the other hand, the differences in the ossification specimens, regardless of whether they were adults or degree of the skull and sometimes of other skeletal ele- juveniles, skeletally mature, or immature, could not ments (e.g., carpals, scapulocoracoids) as well could refer have resulted in cranial kinesis. Thus, if there indeed to different ontogenetic strategies along the diverse evo- was a significant difference in the ossification degree lutionary lineages of pterosaurs: whereas bifenestratan and hereby in the mechanical behavior of the skull and archaeopterodactyloid pterosaurs could have grown between these groups, this difference certainly did not a lifetime long with an accelerated development until lie in the potential of the more basal pterosaurs for cra- reaching sexual maturity and after that a much more nial kinesis. decelerated but still continuous growth (like modern On the other hand, the incomplete fusion state of the crocodiles), derived pterodactyloids could have had deter- skull could have been an ancestral heritage: although minate growth strategy. In this case, continuous growth we know almost nothing about the origin of pterosaurs, would suggest the presence of open sutures a lifetime it is almost certain that the ancestor must be envisioned long, whereas derived pterodactyloids with determinate as a small, arboreal, lizard-like diapsid reptile (Bennett, growth would have completely fused bones after reach- 1997) which was most probably insectivorous (O†si, ing final body size. Bennett (1993) came to a very simi- 2010). Since insectivory has been related to the evolution lar conclusion by suggesting indeterminate growth of cranial kinesis in amniotes (Bout and Zweers, 2001), strategy for bifenestratan and more basal pterodactyloid it is conceivable that the predecessors of pterosaurs pterosaurs after discussing characters that refer to dif- indeed had kinetic skulls, which subsequently became ferent ontogenetic stages in pterosaurs. However, this rigid due to other, yet unknown mechanical constrains. interpretation contradicts to the results of the histologi- Thus the unfused nature and other morphological fea- cal investigations on pterosaur bones which suggest that tures indicative of intracranial movements in the skull pterosaurs, even the more basal ones, had relatively fast of basal pterosaurs can be considered as a residuum of a and deterministic growth (de Ricqle`s et al., 2000; Padian real kinetic skull which is already out of order. In this et al., 2004; Steel, 2008; Chinsamy et al., 2008). case, we might face the phenomenon called transfer Another idea which might give an explanation for the exaptation which refers to the loss of original function of differences includes biomechanical considerations. a certain feature (Arnold, 1994). 828 PRONDVAI AND O†SI CONCLUSIONS Arnold E. 1994. Investigating the origins of performance advantage: adaptation, exaptation and lineage effects. In: Eggleton P, Vane- Our investigation shows that, based on morphological, Wright RI, editors. Phylogeny and ecology. London: Academic comparative anatomical, phylogenetic, and ontogenetic Press. p 123–168. considerations, the skull of most ‘‘adult’’ bifenestratan pter- Arnold E. 1998. Cranial kinesis in lizards: variations, uses, and ori- osaurs (e.g., Eudimorphodon ranzii, Dorygnathus banthen- gins. In: Hecht M, Macintyre R, Clegg M, editors. Evolutionary sis, etc.) and ‘‘juvenile’’ pterodactyloid monofenestratan biology. New York: Plenum Press. Vol 30: p 323–357. specimens (e.g., Tapejara wellnhoferi SMNK PAL 1137) is Arthaber G. 1919. Studien u¨ ber Flugsaurier auf Grund der partially kinetically competent at best. Accordingly their Bearbeitung des Wiener Exemplares von Dorygnathus banthen- sis Theod. sp. Denkschr Akad Wiss Wien math.-nat Kl 97: skull possessed key synovial joints and probably the neces- 391–464. sary protractor muscles but lacked bony gaps or additional Bahl KN. 1937. Skull of Varanus monitor. Rec Ind Mus 39:133–174. mobile regions which would have permitted intracranial Bennett SC. 1993. The ontogeny of Pteranodon and other ptero- movements. Thus, the skull of bifenestratan pterosaurs as saurs. Paleobiology 19:92–106. well as that of the skeletally immature individuals of Bennett SC. 1995. A statistical study of Rhamphorhynchus from derived pterodactyloids was virtually akinetic. the southern limestone of Germany: year classes of a single large The presence of synovial joints and unfused sutures in species. J Vert Paleont 69:569–580. the cranium of bifenestratan pterosaurs and archaeopter- Bennett SC. 1996a. The phylogenetic position of the Pterosauria odactyloids has most probably phylogenetic roots. Among within the Archosauromorpha. Zool J Linn Soc 118:261–308. Bennett SC. 1996b. Year classes of pterosaurs from the Solnhofen diapsids, there is a strong evidence of the plesiomorph na- Limestone of Germany: taxonomic and systematic implications. J ture of streptostyly, which secondarily became restricted Vert Paleont 16:432–444. to an immobile quadrate-squamosal joint in most archo- Bennett SC. 1997. The arboreal leaping theory of the origin of pter- saur clades. Currently, only birds have demonstrably osaur flight. Hist Biol 12:265–290. streptostylic quadrate among archosaurs, but this is Bennett SC. 2001. The osteology and functional morphology of the almost certainly a secondarily regained trait rather than Late Cretaceous pterosaur Pteranodon. Part II. Functional mor- a retained plesiomorphic character. The vestige of the phology. Pala¨ontographica 260A:113–153. streptostylic arrangement in other archosaurs, including Bennett SC. 2007a. A second specimen of the pterosaur Anurogna- bifenestratan and skeletally immature pterodactyloid thus ammoni. Pala¨ontol Z 81:376–396. Bennett SC. 2007b. A review of the pterosaur Ctenochasma: taxon- pterosaurs, might well have represented cartilaginous omy and ontogeny. N Jb Geol Pala¨ont Abh 245:23–31. regions that permitted and facilitated cranial growth dur- Bout RG, Zweers GA. 2001. The role of cranial kinesis in birds. ing ontogeny and/or ensured that the skull could bear rel- Comp Biochem Physiol 131A:197–205. atively high stress and strain loads. Thus the Buffetaut E, Girgorescue D, Csiki Z. 2002. A new giant Pterosaur morphological features indicative of cranial kinesis in with a robust skull from the latest Cretaceous of Romania. Natur- some pterosaurian taxa might well be related to the phe- wissenschaften 89:180–184. nomenon called transfer exaptation. Equally possible is Bu¨ hler P. 1981. Functional anatomy of the avian jaw apparatus. In: the assumption that these characters represent a func- King AS, McLelland J, editors. Form and function in birds. New tionless residuum of a real kinetic skull possessed by the York: Academic Press. Vol 2. p 439–468. Cai Z, Wei F. 1994. ‘‘Zhejiangopterus linhaiensis gen. et sp. nov. hitherto unknown diapsid ancestor of pterosaurs. (Pterosauria) from the Upper Cretaceous of Linhai, Zhejiang, China’’ (In Chinese). Vertebrat Palasiatic 32:181–194. ACKNOWLEDGMENTS Campos DA, Kellner AWA. 1985. Panorama of the flying reptiles in Brazil and South America (Pterosauria|Pterodactyloidea|Anhan- The authors wish to thank M. Witton (University of gueridae). An Acad Bras Cienc 57:141–142, 453–466. Portsmouth, Portsmouth) and the anonymous Reviewer for Chiappe LM, Norell MA, Clark JM. 1998. The skull of a relative of their useful comments that highly improved the standards the stem-group bird Mononykus. Nature 392:275–278. ofthemanuscript;R.Bo¨ttcher and R. Schoch (Staatliches Chinsamy A, Codorniu´ L, Chiappe L. 2008. Developmental growth Museum fu¨ r Naturkunde, Stuttgart), R. Brocke and T. patterns of the filter-feeder pterosaur, Pterodaustro guinazui. Biol Emmel (Senckenberg Forschungsinstitut und Naturmu- Lett 4:282–285. seum, Frankfurt), O. Rauhut (Bayerische Staatsammlung Colbert EH, Russell DA. 1969. The small Cretaceous dinosaur Dro- fu¨r Pala¨ontologie und Geologie, Mu¨nchen) and E. Frey maeosaurus. Am Mus Nov 2380:1–49. Condon K. 1987. A kinematic analysis of mesokinesis in the Nile (Staatliches Museum fu¨ r Naturkunde, Karlsruhe) for access monitor (Varanus niloticus). Exp Biol 47:73–87. to material and helpful assistance; B. Mueller (Texas Tech Dalla Vecchia FM. 2002. Observations on the non-pterodactyloid University, Texas), M. Lamanna (Carnegie Museum of Nat- pterosaur Jeholopterus ningchengensis from the early Cretaceous uralHistory,Pittsburgh),R.Elgin,andV.Griener(Staat- of northeastern China. Nat Nasc 24:8–27. liches Museum fu¨ r Naturkunde, Karlsruhe) for providing Dalla Vecchia FM. 2009. Anatomy and systematic of the pterosaur pictures of indispensible pterosaur specimens; D. Schwarz- Carniadactylus gen. n. rosenfeldi (Dalla Vecchia, 1995). Riv Ital Wings (Museum fu¨ r Naturkunde, Berlin), G. Zboray, and L. Paleont Stratigr 115:159–188. Maka´di (Eo¨tvo¨sLora´nd University, Budapest) for access to de Ricqle`s A, Padian K, Horner JR, Francillon-Viellot H. 2000. Pale- relevant literature and useful discussions. ohistology of the bones of pterosaurs (Reptilia: Archosauria): anat- omy, ontogeny and biochemical implications. Zool J Linn Soc 129:349–385. Dixon AD. 1997. Prenatal development of the facial skeleton. In: LITERATURE CITED Dixon AD, Hoyte DAN, Ro¨nning O, editors. Fundamentals of cra- Andres B, Ji Q. 2006. A new species of Istiodactylus (Pterosauria, niofacial growth. London: CRC Press LLC. p 60–89. Pterodactyloidea) from the Lower Cretaceous of Liaoning, China. Fastnacht M. 2005. Jaw mechanics of the pterosaur skull construc- J Vert Paleont 26:70–78. tion and the evolution of toothlessness. PhD Thesis. Johannes Gu- Andres B, Ji Q. 2008. A new pterosaur from the Liaoning Province tenberg Universita¨t, Mainz, p 31–32, 180–185. of China, the pylogeny of the Pterodactyloidea, and convergence Frazzetta TH. 1962. A functional consideration of cranial kinesis in in their cervical vertebrae. Palaeontology 51:453–469. lizards. J Morph 111:287–319. CRANIAL KINESIS IN PTEROSAURS 829

Frazzetta TH. 1983. Adaptation and function of cranial kinesis in Norman DB, Weishampel DB. 1985. Ornithopod feeding mecha- reptiles: a time–motion analysis of feeding in alligator lizards. In: nisms: their bearing on the evolution of herbivory. Am Nat Rhodin AGJ, Miyata K editors. Advances in herpetology and evo- 126:151–164. lutionary biology: essays in honor of E.E. Williams. Cambridge: O†si A. 2010. Feeding related characters in basal pterosaurs: impli- Museum of Comparative Zoology Publication. p 222–244. cations for dental function, jaw mechanism and diet. Lethaia, Galton PM. 1974. The ornithischian dinosaur Hypsilophodon form DOI: 10.1111/j.1502-3931.2010.00230.x. the Wealden of the Isle of Wight. Bull Brit Mus Nat His 25:1– O†si A, Prondvai E, Frey E, Pohl B. 2010. New interpretation of the 152. palate of pterosaurs. Anat Rec 293:243–258. Gasparini Z, Ferna´ndez M, de La Fuente M. 2004. A new pterosaur Padian K, Horner JR, de Ricqles A. 2004. Growth in small dino- from the Jurassic of Cuba. Palaeontology 47:919–927. saurs and pterosaurs: the evolution of archosaurian growth strat- Gussekloo SWS, Bout RG. 2005. Cranial kinesis in palaeognathous egies:J Vert Paleont 24:555–571. birds. J Exp Biol 208:3409–3419. Patchell FC, Shine R. 1986. Feeding mechanisms in pygopodid liz- Herrel A, De Vree F. 1999. Kinematics of intraoral transport and ards: how can Lialis swallow such large prey? J Herpet 20:59–64. swallowing in the herbivorous lizard Uromastix acanthinurus.J Rieppel O. 1978. The phylogeny of cranial kinesis in lower verte- Exp Biol 202:1127–1137. brates, with special reference to the Lacertilia. N Jb Geol Pala¨ont Herrel A, De Vree F, Delheusy V, Gans C. 1999. Cranial kinesis in Abh 156:353–370. gekkonid lizards. J Exp Biol 202:3687–3698. Rieppel O. 1993. Patterns of diversity in the reptilian skull. In: Herring S, Teng S. 2000. Strain in the braincase and its sutures Hanken J, Hall BK editors. The skull: patterns of structural and during function. Am J Phys Anthropol 112:575–593. systematic diversity. Chicago: University of Chicago Press. p 344– Holliday CM, Witmer LM. 2007. Archosaur adductor chamber evo- 390. lution: integration of musculoskeletal and topological criteria in Rieppel O, Gronowski RW. 1981. The loss of the lower temporal ar- jaw muscle homology. J Morph 268:457–484. cade in diapsid reptiles. Zool J Linn Soc 72:203–217. Holliday CM, Witmer LM. 2008. Cranial kinesis in dinosaurs: intra- Ryabinin AN. 1948. ‘‘Remarks on a flying reptile from the Jurassic cranial joints, protractor muscles, and their significance for cra- of the Kara-Tau’’ (In Russian). Akad Nauk, Paleont Inst, Trudy, nial evolution and function in diapsids. J Vert Paleont 28:1073– Moscow and Leningrad 15:86–93. 7088. Schwarz D, Frey E, Meyer C. 2007. Novel reconstruction of the ori- Hone DWE, Benton MJ. 2008. Contrasting total-evidence and super- entation of the pectoral girdle in sauropods. Anat Rec 290:32–47. tree methods: the origin of the pterosaurs. Zitteliana B 28:35–60. Sereno PC. 1991. Basal archosaurs: phylogenetic relationships and Hooley RW. 1913. On the skeleton of Ornithodesmus latidens; an functional implications. J Vert Paleont 11(Suppl):1–53. Ornithosaur from the Wealden Shales of Atherfield (Isle of Smith KK, Hylander WL. 1985. Strain gauge measurement of meso- Wight). Q J Geol Soc 96:372–422. kinetic movement in the lizard Varanus exanthematicus. J Exp Iordansky NN. 1989. ‘‘Jaw apparatus of perennibranchiate urode- Biol 114:53–70. lans and some problems of heterochronous evolution,’’ (in Rus- Stecher R. 2008. A new Triassic pterosaur from Switzerland (Cen- sian). Zool Zhur 73:87–99. tral Austroalpin, Grisons), Raeticodactylus filisurensis gen. et sp. Iordansky NN. 1990. Evolution of cranial kinesis in lower tetrapods. nov. Swiss J Geosci 101:185–201. Neth J Zool 40:32–54. Steel L. 2008. The palaeohistology of pterosaur bone: an overview. Kellner AWA. 1989. A new edentate pterosaur of the Lower Creta- Zitteliana B 28:109–125. ceous from the Araripe Basin, Northeast Brazil. An Acad Bras Szenta´gothai J, Re´thelyi M. 2006. ‘‘Functional Anatomy I’’ (In Hun- Cienc 61:439–446. garian). Re´thelyi M editor, 8th ed. Budapest: Medicina Press. p Kellner AWA. 1996. Description of the braincase of two early Creta- 375–399. ceous pterosaurs (Pterodactyloidea) from Brazil. Am Mus Nov Summers AP, Wake MH. 2005. The retroarticular process, streptos- 3175:1–34. tyly and the caecilian jaw closing system. Zoology 108:307–315. Kellner AWA. 2003. Pterosaur phylogeny and comments on the evo- Throckmorton G. 1976. Oral food processing in two herbivorous liz- lutionary history of the group: In: Buffetaut E, Mazin JM, editors. ards, Iguana iguana (Iguanidae) and Uromastix aegypticus (Aga- Evolution and palaeobiology of Pterosaurs. Geological Society Spe- midae). J Morphol 148:363–390. cial Publication, no. 217, p 105–137. Unwin DM. 2003. On the phylogeny and evolutionary history o Kellner AWA, Langston W. 1996. Cranial remains of Quetzalcoatlus pterosaurs. In: Buffetaut E, Mazin JM, editors. Evolution and (Pterosauria, Azhdarchidae) from late Cretaceous sediments of Palaeobiology of Pterosaurs. Geological Society Special Publica- Big Bend National Park, Texas. J Vert Paleont 16:222–231. tion, no. 217, p 139–190. Kellner AWA, Tomida Y. 2000. Description of a new species of Veldmeijer AJ, Meijer HJM, Signore M. 2006. Coloborhynchus from Anhangueridae (Pterodactyloidea) with comments on the ptero- the Lower Cretaceous Santana Formation, Brazil (Pterosauria, saur fauna from the Santana Formation (Aptian -Albian), North- Pterodactyloidea, Anhangueridae); an update. www.PalArch.nl. eastern Brazil. Nation Sci Mus Monogr 17:1–135. Vert Paleont 3:15–29. Lee MSY. 2001. Molecules, morphology, and the monophyly of dia- Versluys J. 1910. Streptostylie bei Dinosauriern, nebst Bemerkun- psid reptiles. Contrib Zool 70:1–22. gen u¨ ber die Verwandtschaft der Vo¨gel und Dinosaurier. Zool Jb Lu¨ J, Gao Y, Xing L, Li Z, Ji Q. 2007. A new species of Huaxiapte- Anat 30:175–260. rus (Pterosauria: Tapejaridae) from the early Cretaceous of west- Versluys J. 1912. Das Streptostylie-Problem und die Bewegungen ern Liaoning, China. Acta Geol Sin 81:683–687. im Scha¨del bei Sauropsiden. Zool Jb Anat(Suppl 15):545–716. Lu¨ J, Unwin DM, Jin X, Liu Y, Ji Q. 2009. Evidence for modular Wang X, Kellner AWA, Jiang S, Meng X. 2009. An unusual long- evolution in a long tailed pterosaur with a pterodactyloid skull. tailed pterosaur with elongated neck from western Liaoning of Proc Roy Soc, doi: 10.1098/rspb.2009. 1603. China. An Acad Bras Cieˆnc 81:793–812. Mazzetta GV, Farina RA, Vizcaino SF. 1998. On the palaeobiology of Wang X, Kellner AWA, Zhou Z, Campos D de A. 2007. A new ptero- the South American horned theropod Carnotaurus sastrei saur (Ctenochasmatidae, Archaeopterodactyloidea) from the Bonaparte. In: Pe´rez-Moreno BP, Holtz TR, Sanz JL, Jr., Moratalla Lower Cretaceous Yixian Formation of China. Cretac Res 28:245– JJ, editors. Aspects of theropod paleobiology. Gaia Special Volume 260. No 15. Lisbon: Museu Nacional de Histo´ria Natural. p 185–192. Wang X, Zhou Z. 2003. A new pterosaur (Pterodactyloidea, Tapejari- Metzger K. 2002. Cranial kinesis in Lepidosaurs: skulls in motion. dae) from the Early Cretaceous Jiufotang Formation of Western In: Aerts P, Aout KD, Herrel A, Van Damme R, editors. Topics in Liaoning, China and its implications for biostratigraphy. Chinese functional and ecological vertebrate vertebrate morphology. Maas- Sci Bull 48:16–23. tricht: Shaker Publishing. p 15–46. Wellnhofer P. 1974. Campylognathoides liasicus (Quenstedt), an Norman DB. 1984. On the cranial morphology and evolution of orni- Upper Jurassic pterosaur from Holzmaden—The Pittsburgh speci- thopod dinosaurs. Symp Zool Soc 52:521–547. men. Ann Carnegie Mus 45:5–34. 830 PRONDVAI AND O†SI

Wellnhofer P. 1978. Handbuch der Pala¨oherpetologie. Teil 19. Ptero- Witmer LM. 1995. The Extant phylogenetic bracket and the impor- sauria. Stuttgart: Gustav Fischer Verlag. tance of reconstructing soft tissues in fossils. In: Thomason JJ, Wellnhofer P, Kellner AWA. 1991. The skull of Tapejara wellnhoferi editor. Functional morphology in vertebrate paleontology. New Kellner (Reptilia: Pterosauria) from the Lower Cretaceous San- York: Cambridge University Press. p 19–33. tana Formation of the Araripe Basin, Northeastern Brazil. Mitt Zusi RL. 1984. A functional and evolutionary analysis of rhynchoki- Bayer Staatsslg Pala¨ont Hist Geol 31:89–106. nesis in birds. Smithson Contrib Zool 395:1–40. Wild R. 1978. Die Flugsaurier (Reptilia, Pterosauria) aus der Oberen Zusi RL. 1993. Patterns of diversity in the avian skull. In: Trias von Cene bei Bergamo, Italien. Boll Soc Paleont It 17:176–256. Hanken J, Hall BK, editors. The skull: patterns of structural Wild R. 1984. Flugsaurier aus der Obertrias von Italien. Naturwis- and systematic diversity. Chicago: University of Chicago Press. senschaften 71:1–11. p 391–437.